[LAD] Re: [PATCH] lash: robustify alsa-sequencer event reception
drobilla at connect.carleton.ca
Wed Aug 22 15:21:56 UTC 2007
On Tue, 2007-21-08 at 21:29 +0200, Karsten Wiese wrote:
> trying lash-0.5.3 here, weired things happened:
> I set up a simple project with only 1 lash_synth, the lash_synth
> was connected to an alsa-sequencer keyboard port and to jack.
> After saving, closing and restarting that lash-project,
> the alsa-sequencer connection wasn't restored, until I launched an
> innocent "aconnect -il" in an other terminal.
> Funny, no?
> Reason is I think, that after polling for alsa-seq events, always
> only 1 event was red via snd_seq_event_input().
> Attached patch cures & simplifies things by avoiding the poll() call
> and using pthread_cancel() to stop the thread.
Looks good (at face value anyway...).
Applied to CVS, completely untested. I'll take your word for it ;)
More information about the Linux-audio-dev